Impact of selective decontamination of the digestive tract on carriage and infection due to Gram-negative and Gram-positive bacteria: a systematic review of randomised controlled trials.

SUMMARY

Mete-analyses of randomised controlled trials of selective digestive decontamination have clinical outcome measures, mainly pneumonia and mortality. This mete-analysis has a microbiological endpoint and explores the impact of selective digestive decontamination on Gram-negative and Gram-positive carriage and severe infections We searched electronic databases Cochrane Register of Controlled Trials, previous mete-analyses and conference proceedings with no language restrictions. We included randomised controlled trials which compared the selective digestive decontamination protocol with no treatment or placebo. Three reviewers independently applied selection criteria, performed the quality assessment and extracted the data. The outcome measures were carriage and severe infection due to Gram-negative and Gram-positive bacteria. Odds ratios were pooled with the random effect model. Fifty-four randomised controlled trials comprising 9473 patients were included, 4672 patients received selective digestive decontamination and 4801 were controls Selective digestive decontamination significantly reduced oropharyngeal carriage (odds ratio [OR] 0.13, 95% confidence interval [CI] 0.07 to 0.23), rectal carriage (OR 0.15, 95% CI 0.07 to 0.31), overall infection (OR 0.17, 95% CI 0.10 to 0.28), lower respiratory tract infection (OR 0.11, 95% CI 0.06 to 0.20) and bloodstream infection (OR 0.35, 95% CI 0.21 to 0.67) due to Gram-negative bacteria. Reduction in Gram positive carriage was not significant. Gram-positive lower airway infections were significantly reduced (OR 0.52, 95% CI 0.34 to 0.78). Gram-positive bloodstream infections were not significantly increased (OR 1.03, 95% CI 0.75 to 1.41). The association of parenteral and enteral antimicrobials was superior to enteral antimicrobials in reducing carriage and severe infections due to Gram-negative bacteria. This mete-analysis confirms that selective digestive decontamination mainly targets Gram-negative bacteria, it does not show a significant increase in Gram-positive infection.

Key Words: selective decontamination, digestive decontamination, intensive care unit, respiratory tract infection, carriage, bloodstream infection, Gram-negative, Gram-positive

The key to infection control in the intensive care unit (ICU) is to appreciate that a limited range of potentially pathogenic micro-organisms, both 'normal', including Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus; and 'abnormal', such as aerobic Gram-negative bacilli, may cause three different types of infection, each requiring a different prophylactic manoeuvre (1). Exogenous infections may be controlled by a high level of hygiene, primary endogenous infections by the immediate administration of parenteral antibiotics and secondary endogenous infections by the application of enteral antimicrobials in throat and gut (2).

Selective decontamination of the digestive tract (SDD) using hygiene and parenteral and enteral antimicrobials is a prophylactic measure aiming at the control of exogenous, primary endogenous and secondary endogenous infections, and at the reduction in mortality (1,2). Twelve systematic reviews and mete-analyses of randomised controlled trials (RCT) explored the efficacy of the prophylactic manoeuvre of SDD (3-14). The majority of them focused on a clinical endpoint, mainly pneumonia (3-10,12), bloodstream infection (7,10,14) and mortality (3-9,11,12,14). SDD significantly reduced the odds ratio for pneumonia, bloodstream infection and mortality to 0.35 (95% CI 0.29 to 0.41) (12), 0.63 (0.46 to 0.87) (14) and 0.78 (0.68 to 0.89) (12) respectively. However, none of those meta-analyses assessed the impact on the microbiology of the digestive tract carriage and only two distinguished the type of micro-organism causing infections, one included a small sample of liver transplant patients (11) and the other evaluated only patients with bloodstream infection (14). Additionally, microbiological data are required to confirm or refute the fear that the widespread use of SDD will lead to a serious Gram-positive problem.

We performed a systematic review and meta-analysis of randomised controlled trials of SDD to explore the impact of SDD on carriage and severe infections due to Gram-negative and Gram-positive bacteria.

MATERIALS AND METHODS

Identification of relevant literature and retrieval of studies

We searched Medline (January 1976 to June 2006), Embase (January 1980 to June 2006) and the Cochrane Register of Controlled Trials (June 2006). We use the search terms "intensive care unit", "critical care", "antibiotic combined therapeutic use", "antibiotic combined administration and dosages", "decontamination", "respiratory tract infection prevention and control", "bacterial infection", with the keywords "SDD", "selective decontamination", "selective digestive decontamination", "digestive decontamination", "bowel decontamination". No language restriction was applied. Additionally, we checked reference lists of previous systematic reviews and meta-analyses of SDD and all identified papers of SDD, searched conference abstracts and proceedings of scientific meetings held on the subject.

Inclusion and exclusion criteria 

Inclusion and exclusion criteria were established by the investigators before reviewing abstracts and articles. We included all randomised trials comparing enteral administration of antibiotics of SDD (oropharyngeal, intestinal or both), with or without a parenteral component, with no treatment or placebo in the controls. All published and unpublished trials in unselected and selected critically ill patients were considered. RCTs with usable information by outcome were finally included in the meta-analysis. Studies were excluded for the following reasons: 1) studies including neutropaenic, stem cell and bone marrow transplant patients; 2) non-randomised studies; 3) double publications; 4) studies including data extracted from or complementing main publications; 5) both study arms received SDD but evaluated another drug; and 6) endpoint not infection.

Data extraction and quality assessment

Three investigators (LS, HKFvS, AC) independently retrieved the published findings from each study and compared the sets of data. Any disagreement was resolved by reinspection of the original data and discussion. Where data were available all randomised patients were included in the analysis, allowing an intention-to-treat analysis. The following data were sought for each study: type of population included; specific antimicrobials used and routes; number of patients in each arm; total number of carriers; number of patients with oropharyngeal and rectal carriage due to Gram-negative bacteria; number of patients with Gram-negative infection; number of patients with Gram-negative infection of the lower airways and the bloodstream. Identical variables were sought for Gram-positive bacteria.

We assessed the quality of each study according to a predefined list of seven criteria contained in a scoring system ranging from 0 to 14 and derived by Heyland et al (6), and modified (15), and previously described (14). The assessment was made by three investigators (LS, HKFvS, AC) and included randomisation, blinding, patient selection, population description, reproducibility and definitions of carriage and infection. A total score was obtained as the sum of the subscores of the three evaluators for each of the seven dimensions.

Statistical analysis

The primary endpoints were overall carrier state, patients with oropharyngeal and rectal carriage due to Gram-negative and Gram-positive bacteria, patients with overall Gram-negative and Gram-positive infections, and patients with infections of the lower airway and the bloodstream due to Gram-negative and Gram-positive bacteria.

A subgroup analysis of primary endpoints was planned a priori. To analyse the effect of SDD on the studied variables, RCTs were grouped according to: 1) type of regimen used (parenteral plus enteral or enteral only); 2) quality of randomisation procedures (adequate or inadequate); 3) blinding of patients and caregivers to allocated treatment (blinded or not-blinded); and 4) quality of the study (high or low) (16). Randomisation was adequate when patients were randomised by telephone or a central office. A study was blinded when caregivers and outcome assessors were blinded. The quality categories were obtained according to the median value of the quality scores of all studies. Moreover, an additional subgroup analysis of only infectious endpoints due to Gram-negative micro-organisms was specified a priori in order to analyse the impact of SDD in studies where successful decontamination was achieved. A successful decontamination was defined when the odds ratio (including the 95% confidence interval) for oropharyngeal or rectal or overall carriage due to Gram-negative micro-organisms was less than the unit in each study.

Results were presented as odds ratios (OR) with 95% confidence interval (CI) using the random effects model. The random effects model provides a more conservative estimate of the 95% CI, taking heterogeneity into account: 0.5 cases were added to empty cells to allow calculation of ORs. The ORs were less than the unit if the outcome occurred less frequently in the SDD group. The Cochrane Q statistic for heterogeneity was used both for the outcome measures and through subgroup analyses; we considered heterogeneity to be significant if the P value was < 0.10. We also evaluated the [I.sup.2] measure of inconsistency with the formula 100% x (Q-df)/ Q, where Q is Cochrane's Q statistics and df is the degree of freedom (number of studies-1). Negative values of [I.sup.2] were put equal to 0%, which indicated no observed heterogeneity, while larger percentages indicated increasing heterogeneity. We predefined significant heterogeneity as an [I.sup.2] measure greater than 50% (17). We examined a funnel diagram of the log of the ORs against the weight to estimate potential publication bias. Computations were performed using the EasyMA software (18).

RESULTS

Search findings and general description of the studies

We evaluated 124 potentially eligible studies (Figure 1). Of these studies, 68 were excluded: 47 studies were not randomised, 18 were double publications or included data extracted from the main publication and in three studies both arms received SDD and evaluated another drug. We identified 56 potentially appropriate RCTs. In addition, two studies were excluded because infection or carriage were not the endpoints (19,20). A final sample of 54 RCTs, which enrolled a total of 9473 patients (4672 SDD e 4801 control), was the basis for the systematic review and meta-analysis (21-74).

The details of each study are described in Table 1. One trial was split into two parts in which two different treatments were compared with the same control group (70). Data from the Lingnau's study were retrieved from an additional paper on the microbiology of the same study (75). In one study, one of the two control arms receiving only sucralfate was excluded (51). Of the 54 trials, two were published as abstract (28,37). Four trials were performed in paediatric intensive care units (24,61,64,73). Trials in specific selected types of patients included liver transplantation (23,26,44,56,64,74), burn (24,34), cardia (29,38,73) and gastric surgery (63), oesophageal resection (67), pancreatitis (51), neurosurgery (45,47,69), stroke (42) and acute liver failure (59,60).

[FIGURE 1 OMITTED]

The decontamination protocol varied among studies. Forty-three RCTs included the parenteral component, in general a third generation cephalosporin: 22 in the test arm only and 21 in both arms. In the remaining 11 RCTs including the enteral component only, five used the oropharyngeal and intestinal route, three the intestinal and three the oropharyngeal. Twenty-one RCTs were blinded (23,24,28,29,31,32,34,39,42,43,47,48,50,54,55,57,58,62,63,65,70) and in 16 studies the randomisation was adequate (25,29,31-34,42,47,48,55,57,62-64,73,74). The methodological quality assessment for all trials showed a median of 9.3 (interquartile range 8 to 11) and a weighted kappa on agreement of 0.50 (95% CI, 0.24 to 0.75).

Overall bacterial carriage

Nine studies including 1178 patients (562 SDD, 616 control) reported information on the carrier state, without citing the type of micro-organisms (21,22,24,25,33,49,50,55,58). Ninety-five patients (16.9%) of the SDD group and in 381 patients of the control group (61.8%) developed a carrier state. SDD significantly reduced the number of carriers (OR 0.11, 95% CI 0.05 to 0.26, P < 0.001). The test for heterogeneity yielded a not significant result ([[chi].sup.2] 8.97, P=0.34; [I.sup.2] 10.81%).

Carriage and infection due to Gram-native bacteria

Results from 20 RCTs including 3547 patients (1789 SDD, 1758 control) were available for the analysis of Gram-negative oropharyngeal carriage (21,26,35,36,38,42,43,45-47,50,52,57,61,63,65,68,69,71,73). There were 141 (7.9%) carriers in the SDD group and 536 (30.5%) in the controls. The results indicated a protective effect of SDD on Gram-negative carrier state of the oropharynx (OR 0.13; 95% CI 0.07 to 0.23, P <0.001) (Figure 2). The test for heterogeneity for the overall comparisons was not significant ([[chi].sup.2] 17.69, FL- 0.54; [I.sup.2] 0%).

[FIGURE 2 OMITTED]

Data on rectal carriage were retrieved from 15 RCTs including a total of 1942 patients (971 SDD, 971 control) (23,26,29,30,43,46,47,50,56,57,65,67,68,71,73). There were 69 (7.1%) carriers in the SDD group and 346 (35.6%) amongst controls. SDD significantly reduced Gram-negative rectal carrier state (OR 0.15, 95% CI 0.07 to 0.31, P <0.001). The test for heterogeneity was not significant ([[chi].sup.2] 15.41, P=0.38, [I.sup.2] 2.66%).

Eight studies comprising of 923 patients (451 SDD, 472 control) included data on overall Gram-negative infections with any report of the infection site (21,23,30,33,44,55,62,64). There were 20 (4.4%) patients with Gram-negative infections in SDD group and 89 (18.8%) in the control group. SDD significantly reduced Gram-negative infections by 83% (OR 0.17, 95% CI 0.10 to 0.28, P<0.001) (Figure 3). Heterogeneity was not found ([[chi].sup.2] 5.63, P=0.58; [I.sup.2] 0%).

Fourteen RCTs, including 759 SDD patients and 750 controls, were available for the analysis of Gram-negative lower airway infections (21,22,27,36,45,49,52,53,55,58,64,67,68,73). Twenty-four (3.2%) and 170 (22.7%) patients of the SDD and control group, respectively, developed lower airway infections. The SDD prophylaxis reduced significantly the odds of Gram-negative lower airway infection (OR 0.11, 95% CI 0.06 to 0.20, P< 0.001). The test for heterogeneity was not significant ([[chi].sup.2] 10.13, P=0.68, [I.sup.2] 0%).

[FIGURE 3 OMITTED]

There were 19 trials including 2280 patients (1134 SDD, 1136 controls) which reported data on patients with bloodstream infections due to Gram-negative micro-orgamsms (21,26,27,33,36,40,41,44,45,49,55,59,60,64,66-68,71,73) The prevalence of Gram-negative bloodstream infections was 2% (n=23) among treated patients and 7.7% (n= 87) amongst controls. The results indicated a protective effect of SDD on Gram-negative bloodstream infections (OR 0.35, 95% CI 0.21 to 0.67, P < 0.001). The test for heterogeneity for the overall comparisons was not significant ([[chi].sup.2] 16.65, P=0.65, [I.sup.2] 0%). Results are summarised in Table 2.

Carriage and infection due to Gram-positive bacteria

Table 2 shows the impact of SDD on Gram-positive carriage and infection. SDD reduced, albeit not significantly, oropharyngeal and rectal carriage and overall infections due to Gram-positive bacteria, while lower airway infections were significantly reduced (OR 0.52, 95% CI 0.34 to 0.78, P=0.0016). Gram-positive bloodstream infections were increased by SDD, but not significantly (OR 1.03, 95% CI 0.75 to 1.41, P=0.85). Heterogeneity was not found in all comparisons.

Subgroup analysis

A subgroup analysis was performed in studies including data on patients with Gram-negative oropharyngeal and rectal carriage, and Gram-negative infections, both overall, lower respiratory tract infections (LRTI) and bloodstream infections (BSI) (Table 3). In general, the association of parenteral and enteral antimicrobials was superior to only enteral antimicrobials in reducing oropharyngeal carriage, rectal carriage, overall infections, LRTI and BSI due to Gram-negative bacteria. The reduction in LRTI and BSI was superior in SDD RCTs in which a proper decontamination was obtained.

The subgroup analysis of the primary endpoints due to Gram-positive micro-organisms confirmed the previous pooled data (Table 4). Carriage and overall Gram-positive infections were reduced in the majority of comparisons, but not significantly. Similarly, bloodstream infections were reduced, albeit not significantly, in studies with adequate randomisation, unblinded and with low quality, and in studies using parenteral and enteral anti-microbials. Lower airway infections due to Gram-positive bacteria were reduced in all subgroups.

Effect of publication bias

The inspection of the funnel plots for the outcome variables provided no evidence of publication bias (data not shown).

DISCUSSION

This systematic review of 54 randomised controlled trials assessing selective digestive decontamination in approximately 10,000 patients requiring intensive care is the most comprehensive to date. In particular, this meta-analysis is the first to analyse the microbiology of the SDD-RCTs, both carriage and infection. Four important findings emerge from this meta-analysis:

* SDD significantly reduces both carriage and infections due to Gram-negative bacteria;

* The impact of SDD on Gram-negative carriage and infection using parenteral and enteral antimicrobials is greater than using only enteral antimicrobials;

* The reduction in carriage and infection due to Gram-positive micro-organisms is not significant; lower airway infections due to Gram-positive micro-organisms are significantly reduced, while bloodstream infections due to Gram-positive micro-organisms are not significantly increased;

* The reduction in serious infections is slightly superior in RCTs in which the patients are successfully decontaminated compared with the RCTs in which successful decontamination is not achieved.

This meta-analysis demonstrates that the enteral antimicrobials of SDD, polymyxin and tobramycin, protect against acquisition and secondary carriage due to Gram-negative micro-organisms transmitted via hands of carers. By design of the technique those two antimicrobials were carefully chosen, as they are synergistic against aerobic Gram-negative bacilli, in particular Psaudomonas aeruginosa (76), both respect the indigenous flora of the patients (77,78), neutralise endotoxin released by aerobic Gram-negative bacilli (79), and have a low potential for resistance (80).

Remarkably, the addition of the parenteral antibiotic, mainly cefotaxime, resulted in a more effective clearing of Gram-negative carriage, both oropharyngeal and rectal and reduction in overall Gram-negative infections, LRTI and bloodstream infections compared with RCTs using only enteral antimicrobials. Intravenous cefotaxime is excreted via saliva, bile and mucus into throat and gut, and has been shown to eradicate carriage of 'normal' potential pathogens such as S pneumoniaa S. aureus, H. influenzae and E. coli (81). The greater decontamination effect of SDD using parenteral and enteral antimicrobials may be due to the decontamination of E. coli following cefotaxime excretion in throat and gut.

Data from 56 RCTs and 12 meta-analyses do not provide any evidence for a link between SDD and the emergence of antimicrobial resistance (82). Antimicrobial resistance being a long-term issue, has been evaluated in 11 studies monitoring it between two and nine years, and bacterial resistance associated with SDD has not been a clinical problem (75, 83-92). The experts emphasised that the use of SDD was associated with a Gram-positive problem: however, their claim was mainly based on case reports" and review articles (94,96). This meta-analysis failed to confirm these assertions. Oropharyngeal and rectal carriage of Gram-positive bacteria and overall infections due to Gram-positive bacteria were reduced by SDD, albeit not significantly. More specifically, lower airway infections due to Gram-positive bacteria were significantly reduced, while Gram-positive bloodstream infections were increased, but not significantly. Parenteral cefotaxime given for the first days after admission may effectively control primary endogenous lower airway infections due to the 'community' Gram-positive respiratory pathogens, i.e. S pneumoniae and S. aureus. A substantial part of bloodstream infections are catheter-related and are caused, in general, by skin flora including coagulase-negative staphylococci, which are not influenced by the technique (14). Methicillin-resistant S. aureus (MRSA) and vancomycin-resistant enterococci (VRE) are two Gram-positive bacteria that are intrinsically resistant to the parenteral and enteral antimicrobials of the SDD protocol. They were endemic in the ICUs of nine RCTs (23,34,36,39,43,44,50,70,71) and may explain why the reduction in Gram-positive carriage was not significant. However, VRE carriage and infection were the primary endpoints of SDD RCTs in two American ICUs with endemic VRE (23,44): there was no significant difference between test and control groups. There are seven RCTs conducted in ICUs where MRSA was endemic at the time of the trial, they report a trend towards higher MRSA carriage and infection rates in patients receiving SDD (34,36,39,43,50,70,71). Therefore, the results of this systematic review on Gram-positive micro-organisms should be prudently interpreted, as the impact of SDD could depend on the prevalence or endemicity of Gram-positive organisms in a different study population, irrespective of the effect on Gram-negative micro-organisms.

An intriguing finding of this meta-analysis is the reduced infection rate in RCTs whether the patients were successfully decontaminated or not, i.e. were rendered free of aerobic Gram-negative bacilli or not. This reduction can only be explained by the parenteral component, cefotaxime, that virtually eliminated primary endogenous infections due to normal flora occurring within the first week after admission to the ICU. However, the reduction in infection rate was superior in RCTs in which patients were effectively decontaminated, as there were no secondary endogenous infections in patients rendered free of aerobic Gram-negative bacilli. Similarly, in surgical ICU patients, SDD was beneficial in terms of mortality rate and length of hospital stay only when successful decontamination was achieved (96-98).

We acknowledge some limitations of this review. First, the underreporting of the outcome measures may be explained by the fact that the majority of RCTs of SDD were designed to assess the impact of SDD on lower respiratory tract infections and mortality, not the patient's carrier state and the microbiology of carriage and infections. Second, the design of this review excluded urinary tract infections. Only infections of the lower airways and the bloodstream were included as they contribute to mortality (12,14). Third, the distinction between Gram-negative and Gram-positive micro-organisms was not always obtainable and episodes of infection rather than patients were frequently used in RCTs, making the calculation of the odds ratio impossible. For example, in two RCTs the number of infectious episodes in the control arm exceeded the number of patients enrolled (46,57). Fourth, this review did not include data on the impact of SDD on fungal carriage and infection. Indeed, a previous meta-analysis has demonstrated that SDD significantly reduced both carriage and overall fungal infections, albeit the reduction in fungaemia rate was not significant due to the low fungaemia rate (13). Fifth, by design this review did not distinguish between the type of Gram-negative and Gram-positive microorganism causing infections. This should be taken into account when translating the results of this analysis into clinical practice as mortality is different in severe infections due to the Gram-positive MSSA and MRSA compared with low level pathogens, such as VRE and coagulase-negative staphylococci, and due to the Gram-negative H. influenzae compared with P. aeruginosa.

In summary, this meta-analysis confirms that SDD using parenteral and enteral antimicrobials is a prophylactic protocol that targets mainly Gram-negative micro-organisms. Moreover, the reduction of the level of Gram-negative carriage leads to significantly reduced infection rates. Additionally, the opponents' assertion (99) that there is strong contravening evidence that SDD promotes infection due to Gram-positive bacteria is unsupported by this review.

Accepted for publication on February 6, 2008.

REFERENCES

(1.) van Saene HKF, Petros AJ, Ramsay G, Baxby D. All great truths are iconoclastic: selective decontamination of the digestive tract moves from heresy to level 1 truth. Intensive Care Med 2003; 29:677-690.

(2.) Silvestri L, Mannucci F, van Saene HKF. Selective decontamination of the digestive tract: a life saver. J Hosp Infect 2000; 45:185-190.

(3.) Vandenbroucke-Grauls CMJ, Vandenbroucke JP. Effect of selective decontamination of the digestive tract on respiratory tract infections and mortality in the intensive care unit. Lancet 1991; 338:859-862.

(4.) Selective Decontamination of the Digestive Tract Trialists' Collaborative Group. Meta-analysis of randomised controlled trials of selective decontamination of the digestive tract. BMJ 1993; 307:525-532.

(5.) Kollef M. The role of selective digestive tract decontamination on mortality and respiratory tract infections. A meta-analysis. Chest 1994; 105:1101-1108.

(6.) Heyland DK, Cook DJ, Jaeschke R, Griffith L, Lee HN, Guyatt GH. Selective decontamination of the digestive tract: an overview. Chest 1994; 105:1221-1229.

(7.) Hurley JC. Prophylaxis with enteral antibiotics in ventilated patients: selective decontamination or selective cross-infection? Antimicrob Agents Chemother 1995; 39:941-947.

(8.) D'Amico R, Pifferi S, Leonetti C, Terri V, Tinazzi A, Liberati A. Effectiveness of antibiotic prophylaxis in critically ill adult patients: systematic review of randomised controlled trials. BMJ 1998; 316:1275-1285.

(9.) Nathens AB, Marshall JC. Selective decontamination of the digestive tract in surgical patients. A systematic review of the evidence. Arch Surg 1999; 134:170-176.

(10.) Redman R, Ludington E, Crocker M et al. Analysis of respiratory and non-respiratory infections in published trials of selective digestive decontamination. Intensive Care Med 2001; 27(Suppl 2):S285.

(11.) Safdar N, Said A, Lucey MR. The role of selective digestive decontamination for reducing infection in patients undergoing liver transplantation: a systematic review and meta-analysis. Liver Transpl 2004; 10:817-827.

(12.) Liberati A, D'Amico R, Pifferi S, Leonetti C, Terri V, Brazzi L et al. Antibiotic prophylaxis to reduce respiratory tract infections and mortality in adults receiving intensive care. Cochrane Database Syst Rev 2004; 1:CD000022.

(13.) Silvestri L, van Saene HK, Milanese M, Gregori D. Impact of selective decontamination of the digestive tract on fungal carriage and infection: systematic review of randomised controlled trials. Intensive Care Med 2005; 31:898-910.

(14.) Silvestri L, van Saene HK, Milanese M, Gregori D, Gullo A. Selective decontamination of the digestive tract reduces bacterial bloodstream infections and mortality in critically ill patients. Systematic review of randomized, controlled trials. J Hosp Infect 2007; 65:187-203.

(15.) Brazzi L, Liberati A.. A review of design and conduct of the available studies on selective decontamination of the digestive tract. Reanimation Urgences 1992; 1:501-507.

(16.) Heyland DK, Novak F, Drover JW, Jain M, Su X, Suchner U. Should immunonutrition become routine in critically ill patients? A systematic review of the evidence. JAMA 2001; 286:944-953.

(17.) Higgins JP, Thompson SG, Decks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327:557-560.

(18.) Cucherat M, Boissel JP, Leizorovicz A. EasyMA: a program for the meta-analysis of clinical trials. Comput Methods Programs Biomed 1997; 53:187-190.

(19.) Martinez-Pellus AE, Merino P, Bru M, Conejero R, Seller G, Munoz C et al. Can selective digestive decontamination avoid the endotoxemia and cytokine activation promoted by cardiopulmonary bypass? Crit Care Med 1993; 21:1684-1891.

(20.) Martinez-Pellus AE, Merino P, Bru M, Canovas J, Seller G, Sapina J et al. Endogenous endotoxemia of intestinal origin during cardiopulmonary bypass. Role of type of flow and protective effect of selective digestive decontamination. Intensive Care Med 1997; 23:1251-1257.

(21.) Abele-Horn M, Dauber A, Bauernfeind A, Russwurm W, Seyfarth-Metzger I, Gleich P et al. Decrease in nosocomial pneumonia in ventilated patients by selective oropharyngeal decontamination (SOD). Intensive Care Med 1997; 23:187-195.

(22.) Aerdts SJ, van Dalen R, Clasener HA, Festen J, van Lier HJ, Vollaard EJ. Antibiotic prophylaxis of respiratory tract infection in mechanically ventilated patients. A prospective, blinded, randomized trial of the effect of a novel regimen. Chest 1991; 100:783-791.

(23.) Arnow PM, Carandang GC, Zabner R, Irwin ME. Randomized controlled trial of selective decontamination for prevention of infections following liver transplantation. Clin Infect Dis 1996; 22:997-1003.

(24.) Barret JP, Jeschke MG, Herndon DN. Selective decontamination of the digestive tract on severely burned pediatric patients. Burns 2001; 27:439-445.

(25.) Bergmans DC, Bonten MJ, Gaillard CA, Paling JC, van der Geest S, van Tiel FH et al. Prevention of ventilator-associated pneumonia by oral decontamination. A prospective, randomized, double-blind, placebo-controlled study. Am J Respir Crit Care Med 2001; 164:382-388.

(26.) Bion JF, Badger I, Crosby HA, Hutchings P, Kong KL, Baker J et al. Selective decontamination of the digestive tract reduces Gram-negative pulmonary colonization but not systemic endotoxemia in patients undergoing elective liver transplantation. Crit Care Med 1994; 22:40-49.

(27.) Blair P, Rowlands BJ, Lowry K, Webb H, Armstrong P, Smilie J. Selective decontamination of the digestive tract: a stratified, randomized, prospective study in a mixed intensive care unit. Surgery 1991; 110:303-310.

(28.) Boland JP, Sadler DL, Stewart W et al. Reduction of nosocomial respiratory tract infections in the multiple trauma patients requiring mechanical ventilation by selective parenteral and enteral antisepsis regimen (SPEAR) in the intensive care (abstract). 17th Congress of Chemotherapy, Berlin, 1991;N[degrees]0465.

(29.) Bouter H, Schippers EF, Luelmo SA, Versteegh MI, Ros P, Guiot HF et al. No effect of preoperative selective gut decontamination on endotoxemia and cytokine activation during cardiopulmonary bypass: a randomized, placebo-controlled study. Crit Care Med 2002; 30:38-43.

(30.) Brun-Buisson C, Legrand P, Rauss A, Richard C, Montravers F, Besbes M et al. Intestinal decontamination for control of nosocomial multiresistant gram-negative bacilli. Study of an outbreak in an intensive care unit. Ann Intern Med 1989; 110:873-881.

(31.) Camus C, Bellissant E, Sebille V, Perrotin D, Garo B, Legras A et al. Prevention of acquired infections in intubated patients with the combination of two decontamination regimens. Crit Care Med 2005; 33:307-314.

(32.) Cerra FB, Maddaus MA, Dunn DL, Wells CL, Konstantinides NN, Lehmann SL et al. Selective gut decontamination reduces nosocomial infections and length of stay but not mortality or organ failure in surgical intensive care unit patients. Arch Surg 1992; 127:163-169.

(33.) Cockerill FR 3rd, Muller SR, Anhalt JP, Marsh HM, Farnell MB, Mucha P et al. Prevention of infection in critically ill patients by selective decontamination of the digestive tract. Ann Intern Med 1992; 117:545-553.

(34.) de la cal MA, Cerda E, Garcia-Hierro P, van Saene HK, Gomez-Santos D, Negro E et al. Survival benefit in critically ill burned patients receiving selective decontamination of the digestive tract. A randomized, placebo-controlled, double-blind trial. Ann Surg 2005; 241:424-430.

(35.) de Jonge E, Schultz MJ, Spanjaard L, Bossuyt PM, Vroom MB, Dankert J et al. Effects of selective decontamination of the digestive tract on mortality and acquisition of resistant bacteria in intensive care: a randomised controlled trial. Lancet 2003; 362:1011-1016.

(36.) Ferrer M, Torres A, Gonzalez J, Puig de la Bellacasa J, el-Ebiary M, Roca M et al. Utility of selective decontamination in mechanically ventilated patients. Ann Intern Med 1994; 120:389-395.

(37.) Finch RG, Tomlinson P, Holliday M, et al. Selective decontamination of the digestive tract (SDD) in the prevention of secondary sepsis in a medical/surgical intensive care unit (abstract). 17th Congress of Chemotherapy, Berlin, 1991;N[degrees]0471.

(38.) Flaherty J, Nathan C, Kabins SA, Weinstein RA. Pilot trial of selective decontamination for prevention of bacterial infection in an intensive care unit. J Infect Dis 1990; 162:1393-1397.

(39.) Gastinne H, Wolff M, Delatour F, Faurisson F, Chevret S. A controlled trial in intensive care units of selective decontamination of the digestive tract with nonabsorbable antibiotics. N Engl J Med 1992; 326:594-599.

(40.) Gaussorgues Ph, Salord F, Sirodot M, et al. Efficacite de la decontamination digestive sur la survenue des bacteriemies nosocomiales chez les patients sous ventilation mecanique et recevant des betamimetiques. Rean Soins Intens Med Urg 1991; 7:169-174.

(41.) Georges B, Mazerolles M, Decun J-F et al. Decontamination digestive selective: resultats d'une etude chez le polytraumatise. Reanimation Urgences 1994; 3:621-627.

(42.) Gosney M, Martin MV, Wright AE. The role of selective decontamination of the digestive tract in acute stroke. Age Ageing 2006; 35:42-47.

(43.) Hammond JM, Potgieter PD, Saunders GL, Forder AA. Double-blind study of selective decontamination of the digestive tract in intensive care. Lancet 1992; 340:5-9.

(44.) Hellinger WC, Yao JD, Alvarez S, Blair JE, Cawley JJ, Paya CV et al. A randomized, prospective, double blinded evaluation of selective bowel decontamination in liver transplantation. Transplantation 2002; 73:1904-1909.

(45.) Jacobs S, Foweraker JE, Roberts SE. Effectiveness of selective decontamination of the digestive tract (SDD) in an ICU with a policy encouraging a low gastric pH. Clinical Intensive Care 1992; 3:52-58.

(46.) Kerver AJ, Rommes JH, Mevissen-Verhage EA, Hulstaert PF, Vos A, Verhoef J et al. Prevention of colonization and infection in critically ill patients: a prospective randomized study. Crit Care Med 1988; 16:1087-1093.

(47.) Korinek AM, Laisne MJ, Nicolas MH, Raskine L, Deroin V, Sanson-Lepors MJ. Selective decontamination of the digestive tract in neurosurgical intensive care unit patients: a double-blind, randomized, placebo-controlled study. Crit Care Med 1993;21:1466-1473.

(48.) Krueger WA, Lenhart FP, Neeser G, Ruckdeschel G, Schreckhase H, Eissner HJ et al. Influence of combined intravenous and topical antibiotic prophylaxis on the incidence of infections, organ dysfunctions, and mortality in critically ill surgical patients. A prospective, stratified, randomized, double-blind, placebo-controlled clinical trial. Am J Respir Crit Care Med 2002; 166:1029-1037.

(49.) Laggner AN, Tryba M, Georgopoulos A, Lenz K, Grimm G, Graninger W et al. Oropharyngeal decontamination with gentamycin for long-stay ventilated patients on stress ulcer prophylaxis with sucralfate? Wien Klin Wochenschr 1994; 106:15-19.

(50.) Lingnau W, Berger J, Javorsky F, Lejeune P, Mutz N, Benzer H. Selective intestinal decontamination in multiple trauma patients: prospective, controlled trial. J Trauma 1997; 42:687-694.

(51.) Luiten EJ, Hop WC, Lange JF, Bruining HA. Controlled clinical trial of selective decontamination for the treatment of severe acute pancreatitis. Ann Surg 1995; 222:57-65.

(52.) Palomar M, Alvarez-Lerma F, Jorda R et al. Prevention of nosocomial infection in mechanically ventilated patients: selective digestive decontamination versus sucralfate. Clinical Intensive Care 1997; 8:228-235.

(53.) Pneumatikos I, Koulouras V, Nathanail C, Goe D, Nakos G. Selective decontamination of subglottic area in mechanically ventilated patients with multiple trauma. Intensive Care Med 2002; 28:432-437.

(54.) Pugin J, Auckenthaler R, Lew DP, Suter PM. Oropharyngeal decontamination decreases incidence of ventilator-associated pneumonia. A randomized, placebo-controlled, double-blind clinical trial. JAMA 1991; 265:2704-2710.

(55.) Quinio B, Albanese J, Bues-Charbit M, Viviand X, Martin C. Selective decontamination of the digestive tract in multiple trauma patients. A prospective double-blind, randomized, placebo-controlled study. Chest 1996; 109:765-772.

(56.) Rayes N, Hansen S, Seehofer D, Muller AR, Serke S, Bengmark S et al. Early enteral supply of Lactobacillus and fiber versus selective bowel decontamination: a controlled trial in liver transplant recipients. Transplantation 2002; 74:123-128.

(57.) Rocha LA, Martin MJ, Pita S, Paz J, Seco C, Margusino L et al. Prevention of nosocomial infection in critically ill patients by selective decontamination of the digestive tract. A randomized, double blind, placebo-controlled study. Intensive Care Med 1992; 18:398-404.

(58.) Rodriguez-Roldan JM, Altuna-Cuesta A, Lopez A, Carrillo A, Garcia J, Leon J et al. Prevention of nosocomial lung infection in ventilated patients: use of an antimicrobial pharyngeal nonabsorbable paste. Crit Care Med 1990; 18:1239-1242.

(59.) Rolando N, Gimson A, Wade J, Philpott-Howard J, Casewell M, Williams R. Prospective controlled trial of selective parenteral and enteral antimicrobial regimen in fulminant liver failure. Hepatology 1993; 17:196-201.

(60.) Rolando N, Wade JJ, Stangou A, Gimson AE, Wendon J, Philpott-Howard J et al. Prospective study comparing the efficacy of prophylactic parenteral antimicrobials, with or without enteral decontamination, in patients with acute liver failure. Liver Transpl Surg 1996; 2:8-13.

(61.) Ruza F, Alvarado F, Herruzo R, Delgado MA, Garcia S, Dorao P et al. Prevention of nosocomial infection in a pediatric intensive care unit (PICU) through the use of selective digestive decontamination. Fur J Epidemiol 1998; 14:719-727.

(62.) Sanchez Garcia M, Cambronero Galache JA, Lopez Diaz J, Cerda Cerda E, Rubio Blasco J, Gomez Aguinaga MA et al. Effectiveness and cost of selective decontamination of the digestive tract in critically ill intubated patients. A randomized, double-blind, placebo-controlled, multicenter trial. Am J Respir Crit Care Med 1998; 158:908-916.

(63.) Schardey HM, Joosten U, Finke U, Staubach KH, Schauer R, Heiss A et al. The prevention of anastomotic leakage after total gastrectomy with local decontamination. A prospective, randomized, double-blind, placebo-controlled, multicenter trial. Ann Surg 1997; 225:172-180.

(64.) Smith SD, Jackson RJ, Hannakan CJ, Wadowsky RM, Tzakis AG, Rowe ML. Selective decontamination in pediatric liver transplants. A randomized prospective study. Transplantation 1993; 55:1306-1309.

(65.) Stoutenbeek CF, van Saene HKF, Zandstra DF. Prevention of multiple organ failure by selective decontamination of the digestive tract in multiple trauma patients. In: Faist E, Baue AE, Schildberg FW eds. The immune consequences of trauma, shock and sepsis--mechanisms and therapeutic approach. Lengerich: Pabst Science Publishers 1996. p. 1055-1066.

(66.) Stoutenbeek CF, van Saene HK, Little RA, Whitehead A; Working Group on Selective Decontamination of the Digestive Tract. The effect of selective decontamination of the digestive tract on mortality in multiple trauma patients: a multicenter randomized controlled trial. Intensive Care Med 2007; 33:261-270.

(67.) Tetteroo GW Wagenvoort JH, Castelein A, Tilanus HW, Ince C, Bruining HA. Selective decontamination to reduce gram-negative colonisation and infections after oesophageal resection. Lancet 1990; 335:704-707.

(68.) Ulrich C, Harinck-de Weerd JE, Bakker NC, Jacz K, Doornbos L, de Ridder VA. Selective decontamination of the digestive tract with norfloxacin in the prevention of ICU-acquired infections: a prospective randomized study. Intensive Care Med 1989;15:424-431.

(69.) Unertl K, Ruckdeschel G, Selbmann HK, Jensen U, Forst H, Lenhart FP et al. Prevention of colonization and respiratory infections in long-term ventilated patients by local antimicrobial prophylaxis. Intensive Care Med 1987; 13:106-113.

(70.) Verwaest C, Verhaegen J, Ferdinande P, Schetz M, van den Berghe G, Verbist L et al. Randomized, controlled trial of selective digestive decontamination in 600 mechanically ventilated patients in a multidisciplinary intensive care unit. Crit Care Med 1997; 25:63-71.

(71.) Wiener J, Itokazu G, Nathan C, Kabins SA, Weinstein RA. A randomized, double-blind, placebo-controlled trial of selective digestive decontamination in a medical-surgical intensive care unit. Clin Infect Dis 1995; 20:861-867.

(72.) Winter R, Humphreys H, Pick A, MacGowan AP, Willatts SM, Speller DC. A controlled trial of selective decontamination of the digestive tract in intensive care and its effect on nosocomial infection. J Antimicrob Chemother 1992; 30:73-77.

(73.) Zobel G, Kuttnig M, Grubbauer HM, Semmelrock HJ, Thiel W. Reduction of colonization and infection rate during pediatric intensive care by selective decontamination of the digestive tract. Crit Care Med 1991; 19:1242-1246.

(74.) Zwaveling JH, Mating JK, Klompmaker IJ, Haagsma EB, Bottema JT, Laseur M et al. Selective decontamination of the digestive tract to prevent postoperative infection: a randomized placebo-controlled trial in liver transplant patients. Crit Care Med 2002; 30:1204-1209.

(75.) Lingnau W, Berger J, Javorsky F, Fille M, Allerberger F, Benzer H. Changing bacterial ecology during a five year period of selective intestinal decontamination. J Hosp Infect 1998; 39:195-206.

(76.) Kuipers JS. Combinations of antimicrobial agents. The invitro sensitivity of 100 strains of Pseudomonas aeruginosa to polymyxin B, colistin, carbenicillin, gentamicin and doxycycline and to various combinations of theses antibiotics. Archivum Chirurgicum Neerlandicum 1975; 27:257-270.

(77.) Mulder JG, Wiersma WE, Welling GW, van der Waaij D. Low dose oral tobramycin treatment for selective decontamination of the digestive tract: a study in human volunteers. J Antimicrob Chemother 1984; 13:495-504.

(78.) van Saene JJ, van Saene HK, Tarko-Smit NJ, Beukeveld GJ. Enterobacteriaceae suppression by three different oral doses of polymyxin E in human volunteers. Epidemiol Infect 1988; 100:407-417.

(79.) van Saene JJM, Stoutenbeek CP, van Saene HKF et al. Reduction of the intestinal endotoxin pool by three different SDD regimens in human volunteers. J Endotoxin Res 1996; 3:337-343.

(80.) van Saene HKF, Reilly NJ, de Silvestre A et al. Antibiotic policies in the intensive care unit. In: van Saene HKF, Silvestri L, de la Cal MA, eds. Infection Control in the Intensive Care Unit, 2nd ed. Milan, Italy: Springer Verlag 2005. p. 231-246.

(81.) Ledingham IM, Alcock SR, Eastaway AT, McDonald JC, McKay IC, Ramsay G. Triple regimen of selective decontamination of the digestive tract, systemic cefotaxime, and microbiological surveillance for prevention of acquired infection in intensive care. Lancet 1988; 1:785-790.

(82.) Silvestri L, van Saene HKF. Selective decontamination of the digestive tract does not increase resistance in critically ill patients: Evidence from randomised controlled trials. Crit Care Med 2006; 34:2027-2030.

(83.) Hammond JM, Potgieter PD. Long-term effects of selective decontamination on antimicrobial resistance. Crit Care Med 1995; 23:637-645.

(84.) van der Voort PH, van Roon EN, Kampinga GA, Boerma EC, Gerritsen RT, Egbers PH et al. A before-after study of multi-resistance and cost of selective decontamination of the digestive tract. Infection 2004; 32:271-277.

(85.) Viviani M, van Saene HK, Dezzoni R, Silvestri L, Di Lenarda R, Berlot G et al. Control of imported methicillin-resistant Staphylococcus aureus (MRSA) on mechanically ventilated patients: a dose-response study of enteral vancomycin to reduce absolute carriage and infection. Anaesth Intensive Care 2005; 33:361-372.

(86.) Stoutenbeek CF, van Saene HKF, Zandstra DF. The effect of oral non-absorbable antibiotics on the emergence of resistant bacteria in patients in an intensive care unit. J Antimicrob Chemother 1987; 19:513-520.

(87.) Sarginson RE, Taylor N, Reilly N, Baines PB, van Saene HK. Infection in prolonged pediatric critical illness: A prospective four year study based on knowledge of the carrier state. Crit Care Med 2004; 32:839-847.

(88.) Heininger A, Meyer E, Schwab F, Marschal M, Unertl K, Krueger WA. Effects of long-term routine use of selective digestive decontamination on antimicrobial resistance. Intensive Care Med 2006; 32:1569-1576.

(89.) Leone M, Albanese J, Antonini F, Nguyen-Michel A, Martin C. Long-term (6-year) effect of selective digestive decontamination on antimicrobial resistance in intensive care, multiple-trauma patients. Crit Care Med 2003; 31:2090-2095.

(90.) de la Cal MA, Cerda E, van Saene HK, Garcia-Hierro P, Negro E, Parra ML et al. Effectiveness and safety of enteral vancomycin to control endemicity of methicillin-resistant Staphylococcus aureus in a medical/surgical intensive care unit. J Hosp Infect 2004; 56:175-183.

(91.) Tetteroo GWM, Wagenvoort JHT, Bruining HA. Bacteriology of selective decontamination: efficacy and rebound colonisation. J Antimicrob Chemother 1994; 34:139-148.

(92.) Cerda E, Abella A, de la Cal MA, Lorente JA, Garcia-Hierro P, van Saene HK et al. Enteral vancomycin controls methicillin-resistant Staphylococcus aureus endemicity in an intensive care burn unit: a 9-year prospective study. Ann Surg 2007; 245:397-407.

(93.) Bonten MJ, van Tiel FH, van der Geest S, Stobberingh EE, Gaillard CA. Enterococcus faecalis pneumonia complicating topical antimicrobial prophylaxis. New Engl J Med 1993; 328:209-210.

(94.) Ebner W, Kropec-Hubner A, Daschner FD. Bacterial resistance and overgrowth due to selective decontamination of the digestive tract. Eur J Clin Microbiol Infect Dis 2000; 19:243-247.

(95.) Kollef MIL Selective digestive decontamination should not be routinely employed. Chest 2003; 123 (5 Suppl):464S-468S.

(96.) Tetteroo GW, Wagenvoort JH, Mulder PG, Ince C, Bruining HA. Decreased mortality rate and length of hospital stay in surgical intensive care unit patients with successful decontamination of the gut. Crit Care Med 1993; 21:1692-1698.

(97.) Dive A, Clavero R, Installe E. Reduced mortality rate with gut decontamination. Crit Care Med 1994; 22:1887-1888.

(98.) Stoutenbeek CP, van Saene HKF. Selective decontamination of the digestive tract. In: Pinsky MR, Dahinaut JF, Artigas A, eds. The Splanchnic Circulation. Berlin: Springer 1995. p. 165-174.

(99.) Kallet RH, Quinn TE. The gastrointestinal tract and ventilator-associated pneumonia. Respir Care 2005; 50:910-921.

L. SILVESTRI *, H. K. F. VAN SAENE ([dagger]), A. CASARIN ([double dagger]), G. BERLOT ([section]), A. GULLO ** Department of Emergency, Unit of Anaesthesia and Intensive Care, Presidio Ospedaliero di Gorizia, Gorizia, Italy

* M.D., Head, Department of Emergency, Unit of Anaesthesia and Intensive Care, Presidio Ospedaliero di Gorizia, Gorizia, Italy.

([dagger]) M.D., Ph.D, F.R.C.Path., Consultant/Reader, Department of Medical Microbiology, University of Liverpool and Department of Clinical Microbiology and Infection Control, Alder Hey Children's Hospital, Liverpool, United Kingdom.

([double dagger]) M.D., Clinical Fellow, Department of Critical Care, St. Michael's Hospital, Toronto, Ontario, Canada.

([section]) M.D., Head, Unit of Anesthesia, Intensive Care and Pain Therapy, University Hospital, Trieste, Italy. **M.D., Head, Unit of Anaesthesia and Intensive Care, Policlinico University Hospital, Catania, Italy.

Address for reprints: Dr L. Silvestri, Department of Emergency, Unit of Anaesthesia and Intensive Care, Presidio Ospedaliero, Via Vittorio Veneto 171, 34170 Gorizia, Italy.

TABLE 1
General characteristics of 54 randomised controlled trials of selective
decontamination of the digestive tract

Authors                  Population studied        No. Patients

                                                   SDD       C

Abele-Horn (21)          Trauma                    58        30
Aerdts (22)              Mixed                     28        60
Arnow (23)               Liver transplant          48        38
Barrette (24)            Paediatric, burns         11        12
Bergmans (25)            Mixed                     87        139
Bion (26)                Liver transplant          27        32
Blair (27)               Mixed                     161       170
Boland (28)              Trauma                    32        32
Bouter (29)              Cardiac                   24        27
Brun-Buisson (30)        Mixed                     65        68
Camus (31)               Mixed                     259       256
Cerra (32)               Mixed                     24        23
Cockerill (33)           Mixed                     75        75
de la Cal (34)           Burns                     58        59
de Jonge (35)            Mixed                     466       468
Ferrer (36               Respiratory               51        50
Finch (37)               Mixed                     24        25
Flahertye (38)           Cardiac                   51        56
Gastinne (39)            Mixed                     220       225
Gaussorgues (40)         Mixed                     59        59
Georges (41)             Trauma                    31        33
Gosneyl (42)             Acute stroke              103       100
Hammond (43)             Respiratory               162       160
Hellinger (44)           Liver transplant          37        43
Jacobs (45)              Neurosurgery              45        46
Kerver (46)              Mixed                     49        47
Korinek (47)             Neurosurgery              96        95
Krueger (48)             Surgical, trauma          273       273
Laggner (49)             Mixed                     33        34

Lingnau (50)             Trauma                    180       177

Luiten (51)              Pancreatitis              54        55
Palomar (52)             Trauma                    59        55
Pneumatikos (53)         Trauma                    40        39
Pugin (54)               Surgical, trauma          38        41
Quinio (55)              Trauma                    76        72
Rayes (56)               Liver transplant          32        63
Rochas (57)              Trauma                    47        54
Rodriguez-Roldan (58)    Mixed                     14        17
Rolando (59)             Liver failure             49        52
Rolando (60              Liver failure             47        61

Ruza (61)                Paediatric                116       110
Sanchez-Garcia (62)      Mixed                     131       140
Schardey (63)            Gastrectomy               102       103
Smith (64)               Paediatric, liver         18        18
                         transplant
Stoutenbeek (65)         Trauma                    49        42
Stoutenbeek (66)         Trauma                    201       200
Tetteroo (67)            Esophagectomy             56        58
Ulrich (68)              Mixed                     55        57
Unertl (69)              Neurosurgery              19        20

Verwaest (70)            Mixed                     220       220
                                                   220       220
Wiener72                 Mixed                     30        31
Winter (72)              Mixed                     91        92
Zobel (73)               Paediatric, cardiac       25        25
Zwaveling (74)           Liver transplant          45        44

Authors                                                      Regimen
                        Parenteral
                                                             AGNB

Abele-Horn (21)         Cefotaxime                           P T
Aerdts (22)             Cefotaxime                           P Nor
Arnow (23)              Cefotaxime/ampicillin 2 arms         P G
Barrette (24)           Pip eracillin/amikacin/van 2 arms    P T
Bergmans (25)           Antibiotiotic (40%) 2 arms           P G
Bion (26)               Cefotaxime/ampicillin 2 arms         P T
Blair (27)              Cefotaxime                           P T
Boland (28)             Cefotaxime                           P T
Bouter (29)             Flucloxacillin 2 arms                Pb Neo
Brun-Buisson (30)       -                                    P Neo Nal
Camus (31)              -                                    P T
Cerra (32)              -                                    Nor
Cockerill (33)          Cefotaxime                           Pb G
de la Cal (34)          Cefotaxime                           P T
de Jonge (35)           Cefotaxime                           P T
Ferrer (36              Cefotaxime 2 arms                    P T
Finch (37)              Cefotaxime                           Pb G
Flahertye (38)          Cefazolin 2 arms                     P G
Gastinne (39)           Antibiotics 2 arms (65%)             P T
Gaussorgues (40)        Antibiotics 2 arms                   P G
Georges (41)            Amoxicillin/clavulanate 2 arms       P N
Gosneyl (42)            -                                    P T
Hammond (43)            Cefotaxime 2 arms                    P T
Hellinger (44)          Ceftizoxime 2 arms                   P G
Jacobs (45)             Cefotaxime                           P T
Kerver (46)             Cefotaxime                           P T
Korinek (47)            -                                    P T
Krueger (48)            Ciprofloxacin                        Pb G
Laggner (49)            Amoxicillin/clavulanate (70%)
                        2 arms                               G
Lingnau (50)            Ciprofloxacin 2 arms                 P T
                                                             P Cipro
Luiten (51)             Cefotaxime                           P Nor
Palomar (52)            Cefotaxime                           P T
Pneumatikos (53)        -                                    P T
Pugin (54)              -                                    Pb Neo
Quinio (55)             Cefazolin 2 arms (38%)               P G
Rayes (56)              Ceftriaxone/metronidazole 2 arms     P T
Rochas (57)             Cefotaxime                           P T
Rodriguez-Roldan (58)   -                                    P T(or N)
Rolando (59)            Cefuroxime                           P T
Rolando (60             Ceftazidime/flucloxacillin 2 arms    P T

Ruza (61)               -                                    P T
Sanchez-Garcia (62)     Ceftriaxone                          P G
Schardey (63)           Cefotaxime 2 arms                    Pb T
Smith (64)              Cefotaxime/ampicillin 2 arms         P T

Stoutenbeek (65)        Cefotaxime 2 arms                    P T
Stoutenbeek (66)        Cefotaxime                           P T
Tetteroo (67)           Cefotaxime                           P T
Ulrich (68)             Trimetoprim                          P Nor
Unertl (69)             -                                    Pb G

Verwaest (70)           Cefotaxime                           P T
                        Ofloxacin                            Ofloxacin
Wiener72                -                                    P G
Winter (72)             Ceftazidime                          P T
Zobel (73)              Cefotaxime                           P G
Zwaveling (74)          Cefotaxime/tobramycin 2 arms         P T

Authors                      Enteral

                         Yeasts        MRSA        Site

Abele-Horn (21)          A                         O, -
Aerdts (22)              A                         O,I -
Arnow (23)               Ny                        O,I -
Barrette (24)            A                         -, I
Bergmans (25)            -             Van         O, -
Bion (26)                A                         O,I -
Blair (27)               A                         O,I -
Boland (28)              Ny                        O,I -
Bouter (29)              -                         O, -
Brun-Buisson (30)        -                         -, I
Camus (31)               -                         O,I -
Cerra (32)               Ny                        -, I
Cockerill (33)           Ny                        O,I -
de la Cal (34)           A                         O,I -
de Jonge (35)            A                         O,I -
Ferrer (36               A                         O,I
Finch (37)               A                         O,I
Flahertye (38)           Ny                        O,I
Gastinne (39)            A                         O,I
Gaussorgues (40)         A             Van         -, I
Georges (41)             A                         O,I
Gosneyl (42)             A                         O,I
Hammond (43)             A                         O,I
Hellinger (44)           Ny                        1
Jacobs (45)              A                         O,I
Kerver (46)              A                         O,I
Korinek (47)             A             Van         O,I
Krueger (48)             -             Van         1
Laggner (49)             A                         O'_
                         2 arms
Lingnau (50)             A                         O,I
                         A                         O,I
Luiten (51)              A                         O,I
Palomar (52)             A                         1
Pneumatikos (53)         A                         O'_
Pugin (54)               -             Van         O'_
Quinio (55)              A                         O,I
Rayes (56)               A                         -, I
Rochas (57)              A                         O,I
Rodriguez-Roldan (58)    A                         O'_
Rolando (59)             A, Clo        Mup         O,I
Rolando (60              A                         O,I
                         2 arms
Ruza (61)                Ny                        -, I
Sanchez-Garcia (62)      A                         O,I
Schardey (63)            A             Van         - ,I
Smith (64)               A                         O,I

Stoutenbeek (65)         A                         O,I
Stoutenbeek (66)         A                         O,I
Tetteroo (67)            A                         O,I
Ulrich (68)              A                         O,I
Unertl (69)              A                         1
                         (only O)
Verwaest (70)            A                         O,I
                                                   O,I
Wiener72                 Ny                        O,I
Winter (72)              A                         O,I
Zobel (73)               A                         O,I
Zwaveling (74)           A                         O,I

AGNB=aerobic Gram-negative bacilli, MRSA=methicillin-resistant,
Scaphylococcus aureus, SDD=selective decontamination of the
digestive tract, C=control, A=amphotericin B, Cipro=ciprofloxacin,
Cie=clotrimazole, G=gentamicin, Mup= mupirocin, Nat=nalidixic
acid, Neo=neomycin, N=netilmicin, Nor=norfloxacin, Ny=nystatin,
P=polymyxin E, Pb=polymyxin B, T=tobramycin, Van=vancomycin,
O=oropharynx, I=intestine, MU=million units.

TABLE 2
Meta-analysis of RCTs on the effect of SDD on Gram-negative and
Gram-positive bacterial carriage and infections

Endpoints                    No. RCTs     No. patients      No. events
                                          SDD       C       SDD     C
Gram-negative

 Oropharyngeal carriage      20           1789      1758    141     536

 Rectal carriage             15           971       971     69      346

 Overall infections          8            451       472     20      89

 LRTI                        14           759       750     24      170

 Bloodstream infection       19           1134      1136    23      87

Gram-positive

 Oropharyngeal carriage      12           1129      1093    62      110

 Rectal carriage             6            410       403     42      64

 Overall infections          4            266       252     25      26

 LRTI                        14           585       593     49      80

 Bloodstream infectons       19           1134      1146    104     103

Endpoints                    OR (95% CI)        P

Gram-negative

 Oropharyngeal carriage      0.13 (0.07-0.23)   <0.001

 Rectal carriage             0.15 (0.07-0.31)   <0.001

 Overall infections          0.17 (0.10-0.28)   <0.01

 LRTI                        0.11 (0.06-0.20)   <0.001

 Bloodstream infection       0.35 (0.21-0.67)   <0.001

Gram-positive

 Oropharyngeal carriage      0.55 (0.30-1.02)   0.06

 Rectal carriage             0.53 (0.12-2.43)   0.41

 Overall infections          0.76 (0.41-1.40)   0.30

 LRTI                        0.52 (0.34-0.78)   0.0016

 Bloodstream infectons       1.03 (0.75-1.41)   0.85

RCTs=randomised controlled trials, SDD= selective decontamination
of the digestive tract, C= controls, OR= odds ratio,
CI=confidence interval, LRTI= lower respiratory tract infection.
The Q and [I.sup.2] tests for heterogeneity were not significant
in all comparisons.

TABLE 3

Subgroup analysis of carriage and infections due to Gram-negative
bacteria

Endpoints                           No. RCTs    No. patients
                                                SDD       C

Oropharyngeal carriage

 Parenteral plus enteral            15          1425      1402

 Enteral only                       5           395       387

 Randomisation adequate             5           372       377

 Randomisation inadequate           15          1417      1381

 Blinded                            7           738       731

 Not blinded                        13          1051      1027

 High quality                       13          1323      1280

 Low quality                        7           466       470

Rectal carriage

 Parenteral plus enteral            10          718       709

 Enteral only                       5           253       262

 Randomisation adequate             4           212       220

 Randomisation inadequate           11          759       751

 Blinded                            7           596       596

 Not blinded                        8           375       375

 High quality                       7           467       467

 Low quality                        8           504       504

Overall infections

 Parenteral plus enteral            6           367       344

 Enteral only                       2           105       107

 Randomisation adequate             3           224       223

 Randomisation inadequate           5           248       218

 Blinded                            1           131       140

 Not blinded                        7           341       311

 High quality                       2           189       170

 Low quality                        6           283       281

 Successfully decontaminated        2           106       68

 Not successfully decontaminated    2           140       143

LRTI

 Parenteral plus enteral            11          609       622

 Enteral only                       3           130       128

 Randomisation adequate             3           139       134

 Randomisation inadequate           11          600       616

 Blinded                            2           90        89

 Not blinded                        12          649       661

 High quality                       5           481       494

 Low quality                        9           258       256

 Successfully decontaminated        9           447       451

 Not successfully decontaminated    2           73        72

Bloodstream infections

 Parenteral plus enteral            17          1028      1043

 Enteral only                       2           106       103

 Randomisation adequate             4           194       190

 Randomisation inadequate           15          940       996

 Blinded                            1           76        72

 Not blinded                        18          1058      1074

 High quality                       7           304       275

 Low quality                        12          830       871

 Successfully decontaminated        9           474       447

 Not successfully decontaminated    2           57        63

Endpoints                     No. events      OR (95% CI)        P
                              SDD     C

Oropharyngeal carriage

 Parenteral plus enteral       110    464    0.09 (0.04-0.18)   <0.001

 Enteral only                  27     73     0.39 (0.19-0.81)   0.011

 Randomisation adequate        21     88     0.21 (0.08-0.60)   0.004

 Randomisation inadequate      120    448    0.11 80.05-0.21)   <0.001

 Blinded                       41     228    0.14 (0.06-0.35)   <0.001

 Not blinded                   100    308    0.12 (0.06-0.25)   <0.001

 High quality                  110    338    0.15 (0.08-0.28)   <0.001

 Low quality                   31     198    0.11 (0.06-0.19    <0.001

Rectal carriage

 Parenteral plus enteral       50     297    0.11 (0.53-0.24)   <0.001

 Enteral only                  19     49     0.26 (0.05-1.35)   0.11

 Randomisation adequate        21     61     0.19 (0.04-1.03)   0.054

 Randomisation inadequate      48     286    0.13 (0.06-0.29)   <0.001

 Blinded                       53     231    0.16 (0.06-0.44)   <0.001

 Not blinded                   16     115    0.12 (0.03-0.43)   <0.001

 High quality                  30     101    0.31 (0.11-0.87)   0.026

 Low quality                   39     245    0.08 (0.04-0.16)   <0.001

Overall infections

 Parenteral plus enteral       18     78     0.15 (0.09-0.27)   <0.001

 Enteral only                  2      11     0.19 (0.05-0.83)   0.027

 Randomisation adequate        11     45     0.19 (0.09-0.39)   <0.001

 Randomisation inadequate      9      44     0.13 (0.06-0.30)   <0.001

 Blinded                       5      19     0.25 (0.09-0.26)   NE

 Not blinded                   15     70     0.14 (0.08-0.26)   <0.001

 High quality                  7      35     0.10 (0.01-0.75)   0.025

 Low quality                   13     54     0.18 (0.10-0.35)   <0.001

 Successfully decontaminated   6      9      0.08 (0.01-0.43)   0.003

 Not successfully
  decontaminated               3      18     0.16 (0.05-0.54)   0.003

LRTI

 Parenteral plus enteral       9      127    0.07 (0.04-0.13)   <0.001

 Enteral only                  15     43     0.28 (0.11-0.68)   0.005

 Randomisation adequate        13     33     0.33 (0.16-0.70)   = 0.004

 Randomisation inadequate      11     137    0.08 (0.04-0.14)   <0.001

 Blinded                       13     35     0.15 (0.01-2.45)   0.18

 Not blinded                   11     135    0.08 (0.04-0.14)   <0.001

 High quality                  8      81     0.11 (0.05-0.21)   <0.001

 Low quality                   16     89     0.07 (0.02-0.37)   <0.001

 Successfully decontaminated   19     124    0.08 (0.03-0.21)   <0.001

 Not successfully
  decontaminated               1      16     0.09 (0.02-0.54)   0.008

 Bloodstream infections

 Parenteral plus enteral       23     80     0.36 (0.22-0.60)   -0.001

 Enteral only                  0      7      0.08 (0.01-1.48)   0.089 *

 Randomisation adequate        0      19     0.05 (0.01-0.45)   0.007

 Randomisation inadequate      23     68     0.39 (0.23-0.64)   -0.001

 Blinded                       0      6      0.03 (0.01-1.92)   NE

 Not blinded                   23     81     0.36 (0.22-0.59)   -0.001

 High quality                  6      15     0.61 (0.23-1.67)   0.34

 Low quality                   17     72     0.29 (0.15-0.54)   -0.001

 Successfully decontaminated   5      29     0.35 (0.12-0.98)   0.045

 Not successfully
  decontaminated               2      1      0.03 (0.03-83.9)   0.82

RCTs=randomised controlled trials, SDD=selective decontamination of the
digestive tract, C= control, OR=odds ratio, CI= confidence interval,
LRTI=lower airway infection. NE=not evaluated as only one study was
included. The Q and h tests for heterogeneity were not significant
in all comparisons except for * where [I.sup.2] was greater than 50%.

TABLE 4
Subgroup analysis of carriage and infections due to Gram-positive
bacteria

Endpoints                         No. RCTs     No. patients
                                               SDD       C
Oropharyngeal carriage

 Parenteral plus enteral          8            868       837

 Enteral only                     4            261       256

 Randomisation adequate           2            199       198

 Randomisation inadequate         10           930       895

 Blinded                          3            248       240

 Not blinded                      9            881       853

 High quality                     9            988       949

 Low quality                      3            141       144

Rectal carriage

 Parenteral plus enteral          3            160       157

 Enteral only                     3            250       246

 Randomisation adequate           2            145       137

 Randomisation inadequate         4            265       266

 Blinded                          3            183       178

 Not blinded                      3            227       225

 High quality                     4            299       288

 Low quality                      2            111       115

Overall infections

 Parenteral plus enteral          3            226       213

 Enteral only                     1            40        39

 Randomisation adequate           1            131       140

 Randomisation inadequate         3            135       112

 Blinded                          1            131       140

 Not blinded                      3            135       112

 High quality                     2            189       170

 Low quality                      2            77        82

LRTI

 Parenteral plus enteral          11           455       465

 Enteral only                     3            130       128

 Randomisation adequate           3            115       114

 Randomisation inadequate         11           470       479

 Blinded                          2            90        89

 Not blinded                      12           504       495

 High quality                     6            271       271

 Low quality                      8            314       322

Bloodstream infection

 Parenteral plus enteral          17           1028      1043

 Enteral only                     2            106       103

 Randomisation adequate           3            169       165

 Randomisation inadequate         16           965       981

 Blinded                          1            76        72

 Not blinded                      18           1058      1074

 High quality                     7            304       275

 Low quality                      12           830       871

Endpoints                     No. events      OR (95% CI)         P
                              SDD   C

Oropharyngeal carriage        46    82     0.52 (0.25-1.10)     0.088

 Parenteral plus enteral      16    28     0.65 (0.17-2.48)     0.52

 Enteral only                 9     40     0.16 (0.07-0.35)     <0.001

 Randomisation adequate       53    70     0.66 (0.37-1.79)     0.16

 Randomisation inadequate     9     40     0.16 (0.08-0.36)     <0.001

 Blinded                      53    70     0.65 (0.35-1.21)     0.18

 Not blinded                  53    93     0.56 (0.25-1.26)     0.16

 High quality                 9     17     0.50 (0.21-1.18)     0.11

 Low quality

Rectal carriage

 Parenteral plus enteral      42    60     0.72 (0.10-5.30)     0.75

 Enteral only                 0     4      0.25 (0.01-4.15)     0.33

 Randomisation adequate       5     0      7.01 (0.27-185.39)   0.24

 Randomisation inadequate     37    64     0.33 (0.07-1.56)     0.16

 Blinded                      5     4      1.06 (0.03-41.73)    0.98

 Not blinded                  37    60     0.41 (0.07-2.34)     0.32

 High quality                 5     4      1.04 (0.07-15.69)    0.98

 Low quality                  37    60     0.38 (0.05-2.74)     0.33

Overall infections

 Parenteral plus enteral      22    20     0.85 (0.44-1.66)     0.64

 Enteral only                 3     6      0.45 (0.10-1.93)     NE

 Randomisation adequate       4     5      0.85 (0.22-3.24)     NE

 Randomisation inadequate     21    21     0.74 (0.37-1.46)     0.39

 Blinded                      4     5      0.85 (0.22-3.24)     NE

 Not blinded                  21    21     0.74 (0.37-1.46)     0.39

 High quality                 15    13     0.71 (0.31-1.63)     0.42

 Low quality                  10    13     0.81 (0.31-2.10)     0.67

LRTI

 Parenteral plus enteral      24    38     0.59 (0.34-1.02)     0.061

 Enteral only                 25    42     0.59 (0.40-0.89)     0.011

 Randomisation adequate       23    36     0.47 (0.24-0.91)     0.024

 Randomisation inadequate     26    44     0.55 (0.32-0.93)     0.025

 Blinded                      22    36     0.44 (0.23-0.86)     0.017

 Not blinded                  27    44     0.57 (0.34-0.96)     0.033

 High quality                 36    53     0.51 (0.31-0.85)     0.01

 Low quality                  13    27     0.53 (0.27-1.04)     0.066

Bloodstream infection

 Parenteral plus enteral      84    90     0.94 (0.66-1.33)     0.071

 Enteral only                 20    13     1.63 (0.76-3.48)     0.21

 Randomisation adequate       27    28     0.74 (0.22-2.46)     0.62

 Randomisation inadequate     77    75     1.05 (0.74-1.50)     0.77

 Blinded                      13    7      1.65 (0.64-4.26)     NE

 Not blinded                  91    95     0.97 (0.69-1.36)     0.87

 High quality                 28    19     1.36 (0.72-2.56)     0.35

 Low quality                  76    84     0.94 (0.65-1.37)     0.75

RCTs=randomised controlled trial, SDD=selective decontamination of the
digestive tract, C=controls; OR=odds ratio, CI=confidence interval,
LRTI=lower respiratory tract infection, NE=not evaluated as only one
study was included. The Q and I[sup.2] tests for heterogeneity were not
significant in all comparisons.